Model Predictive Control Motion Cueing with Nonlinear Constraints and Vestibular Feedback for Serial Robot Motion Simulators

Motion Cueing Algorithms (MCAs) encode the movement of simulated vehicles into movement that can be reproduced with a motion simulator to provide a realistic driving experience within the capabilities of the machine. Herein, an extension of an existing state-of-the-art Nonlinear Model Predictive Control (NMPC) based MCA for serial robot simulators is presented. The proposed novelties are the use of nonlinear constraints to enforce safety protocols, and the use of vestibular feedback, which has been shown to increase motion cueing quality. In addition, the modified algorithm was implemented as part of the Python library Serial Robot Motion Simulator (pySRMS), a package for motion cueing research. PySRMS offers tools for flexible development of NMPC based MCAs for serial robots such as, exposure of all intermediate variables in the plant model, helper functions to define constraints, and the ability to include or exclude vestibular feedback. The capabilities of the new MCA and pySRMS are showcased via three case studies that explore the use of linear and nonlinear constraints, different nonlinear program solvers and integration schemes, and vestibular feedback, respectively. Results from these simulation experiments show that the inclusion of nonlinear constraints and vestibular feedback significantly increases the safety and performance of the proposed MCA compared to the existing, state-of-the-art version.